Preoperative plasma bilirubin may be associated with postoperative delirium in old patients undergoing knee/hip arthroplasty: the PNDRFAP study and the PNDABLE study

doi:10.21203/rs.3.rs-3916412/v1

Download PDF

Research Article

Preoperative plasma bilirubin may be associated with postoperative delirium in old patients undergoing knee/hip arthroplasty: the PNDRFAP study and the PNDABLE study

https://doi.org/10.21203/rs.3.rs-3916412/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background Along with society'saging, the prevalence rate of postoperative delirium (POD) is increasing. Assessing the risk factors of delirium is essential to diagnose and treatpostoperative delirium (POD). The aim of this study was to examine the correlation between bilirubin levels and postoperative delirium.

Methods This study included 621 patients, of whom were from the Perioperative Neurocognitive Impairment and Biomarker Lifestyle (PNDABLE) database. The remaining 577 patients were from the Perioperative Neurocognitive Impairment Risk Factors and Prognosis (PNDRFAP) study. The Mini-Mental State Examination (MMSE) and Confusion Assessment Method (CAM) were used to assess cognitive function before operation, while Memorial Delirium Assessment Scale (MDAS) was used to measure the severity of postoperative delirium (POD). Cerebrospinal fluid biomarkers, including Aβ42, T-tau, P-tau, Aβ42/T-tau, and Aβ42/P-tau, were measured using enzyme-linked immunosorbent assay (ELISA). We conducted logistic regression using data from patients in the PNDRFAP study and the PNDABLE study，which collectively identified risk factors for postoperative cognitive impairment. Nomograms and ROC curves were used as evaluation indices to assess the predictive effect of bilirubin on postoperative delirium (POD). POD events can be predicted by creating a dynamic nomogram model of the network. Finally, a mediation analysis of 10,000 bootstrap iterations was conducted to assess the mediating role of CSF biomarkers.

Results The logistic regression models of the PNDABLE and PNDRFAP studies showed that total bilirubin, direct bilirubin and indirect bilirubin were higher in the postoperative delirium group than in the non-postoperative delirium group. These results suggest that total bilirubin, direct bilirubin and indirect bilirubin may be risk factors for delirium. [PNDRFAP: Total Bilirubin, OR (95% CI): 1.839 (1.645-2.056), P < 0.001; Direct Bilirubin, OR (95% CI): 1.725 (1.448-2.054), P < 0.001; Indirect Bilirubin, OR (95% CI): 1.558 (1.424-1.705), P < 0.001. PNDABLE: Total Bilirubin, OR (95% CI): 1.203 (1.137-1.274), P < 0.001; Direct Bilirubin, OR (95% CI): 1.724 (1.465-2.030), P < 0.001; Indirect Bilirubin, OR (95% CI): 1.182 (1.109-1.260), P < 0.001]. Adjusted for age, sex, educational level and MMSE, total bilirubin, direct bilirubin and indirect bilirubin remained related to POD. Mediation analysis indicted that T-tau mediated the effect of total bilirubin, direct bilirubin and indirect bilirubin on delirium. The respective proportions are 9.08%, 10.26% and 8.70%.

Conclusions Total bilirubin, direct bilirubin and indirect bilirubin are preoperative risk factors associated with an increased risk of postoperative delirium (POD). The association between bilirubin and postoperative delirium may be partly attributed to T-tau levels in cerebrospinal fluid.

Clinical Trial Registration Clinical Registration No. ChiCTR2000033439.

Bilirubin

postoperative delirium

cerebrospinal fluid

logistic regression

Delirium is a neurocognitive disorder characterized by an acute and temporary disturbance in attention, awareness, and cognition[1]. Postoperative delirium (POD) is a common complication among elderly patients, occurring within 3 to 5 days after surgery[2]. Moreover, the incidence rate of high-risk patients, such as elderly patients with prior cognitive dysfunction, is also high[3]. Postoperative delirium has been shown to be strongly associated with higher mortality rates. It not only prolongs the postoperative hospital stay, leading to long-term cognitive impairment after surgery and causing postoperative adverse outcomes, but also places a heavy burden on healthcare resources[4, 5]. Therefore, the identification of high risk for postoperative delirium (POD) in clinical diagnosis and therapy is essential.

At present, several risk factors have been suggested for postoperative delirium, including amyloid β 42 (Aβ42), total tau (T-tau), and phospho-tau (P-tau) [6, 7] T-tau plays an important role in neuronal plasticity by binding tubulin to maintain microtubule stability and induce microtubule formation. Abnormally phosphorylated T-tau protein is deposited in cells, forming neuronal fiber tangles that lead to cognitive dysfunction[8]. Amyloid β 42, a member of the amyloid family, is a polypeptide molecule formed through the proteolytic cleavage of amyloid β precursor proteins. Aβ42 deposition leads to the formation of age spots and exerts strong neurotoxic effects, causing neuronal degeneration and death, which in turn affects cognitive function[9, 10]. Moreover, cerebrospinal fluid (CSF) amyloid β 42 (Aβ42), total tau (T-tau), and phospho-tau (P-tau) are established as biomarkers of postoperative delirium, reflecting plaque pathology, neurodegeneration, and neurofibrillary tangle pathology [11]. Numerous factors contribute significantly to metabolic disorders of the central nervous system, and various markers in the cerebrospinal fluid (CSF) may indicate dysfunction in the central nervous system.

Bilirubin, the primary metabolite of heme compounds in the body, is generally considered potentially cytotoxic and can cause irreversible damage to the brain and nervous system. However, it also has antioxidant functions that can inhibit the oxidation of linoleic acid and phospholipids [12–16]. The association between bilirubin and postoperative delirium has yielded varying results in different studies. In recent studies, the difference in total bilirubin levels between patients with delirium and those without delirium was found to be not significant [17–19]. However, it was found that the level of total bilirubin was obviously increased in delirium patients compared in delirium patients compared with those without delirium [20]. Past research has demonstrated that increased serum bilirubin levels within the normal range can effectively protect against neurological damage following oxidative stress [21]. Abnormal levels of bilirubin in the brain have been reported to cause activation of microglia and astrocytes, impaired myelination, and neuronal cell death [22, 23]. Therefore, we question the validity of the current research results and reinvestigate the correlation between bilirubin and postoperative delirium.

In this study, we aimed to explore bilirubin level in patients with POD and their relationship with CSF Alzheimer-related biomarkers, so as to determine whether high bilirubin is a potential risk factor for POD. We designed the following steps: (1) The difference in uric acid levels between POD and non-POD patients was investigated to evaluate whether high bilirubin is a risk factor for POD;(2) To investigate the association between bilirubin and POD

PNDRAFP

Study registration and data availability

The Perioperative Neurocognitive Disorder Risk Factor and Prognosis (PNDRFAP) study is a large-scale group trial designed to explore the risk factors and biological markers for perioperative neurocognitive disorders in the Han population in northern China [24, 25]. The PNDRFAP study involved only those with general anesthesia. There was no overlap in patients between the PNDABLE study and the PNDRFAP study. This trial has been registered in the Chinese Clinical Trial Registry (clinical registration number PNDRAFP: ChiCTR2000033639) and was endorsed by Qingdao Municipal Hospital’s Ethics Committee.

Setting and participants

The research was carried out in collaboration with the Neurological Division and Anesthesiology Division of Rehabilitation University at Qingdao Municipal Hospital. A trial of the Perioperative Neurocognitive Disorder Risk Factor and Prognosis (PNDRFAP) study was conducted form Jan. 2020 to May 2023. The criteria for inclusion in PNDRAFP were: (1) patients aged 50 to 90 years; (2) american Society of Anesthesiologists (ASA) I-II; (3) patients whose cognitive function was intact before surgery; and (4) patients who were sufficiently educated to complete preoperative neuropsychological examinations. Exclusion criteria were: (1) Mini Mental State Examination (MMSE) score below 24; and (2) severe mental or hearing impairment.

Anesthesia and surgery

Patients received general anesthesia in PNDRFAP. Anesthesia induction involved the use of sufentanil at a dosage of 0.2–0.5 micrograms/kg, ceftriaxone at a dosage of 0.15–0.2 mg/kg, and etomidate at a dosage of 0.15–0.3 mg/kg. Remifentanil was infused continuously at 0.25–2 micrograms/kg/minute for postoperative analgesia. After induction, Cis-atracurium should be given every 40 minutes and stopped 1 hour prior to completion. Depending on anesthetic depth, 0.5-3% sevoflurane is administered by inhalation.

PNDABLE

Study registration and data availability

The Perioperative Neurocognitive Disorder and Biomarker Lifestyle (PNDABLE) study is an ongoing, large-sample cohort study that focused on identifying risk factors and biomarkers for perioperative neurocognitive dysfunction in the northern Chinese Han population. The study aims to detect, diagnose, and intervene in perioperative neurocognitive disorder (POD) at an earlier stage [26, 27]. The PNDABLE study included a combination of spinal and extradural anesthesia and collected of CSF before surgery for biomarker analysis. The study was registered with Qingdao Municipal Hospital Ethics Committee (Clinical Registration No: ChiCTR2000033439).

Setting and Participants

This study was conducted jointly by the Department of Neurology and Anesthesiology at Qingdao Municipal Hospital, which is affiliated with Rehabilitation University. The Perioperative Neurocognitive Disorder and Biomarker Lifestyle (PNDABLE) study will take place from November 2020 to August 2023.The criteria for inclusion in PNDABLE were: (1) patients aged 50 to 90 years; (2) ASA I-II; (3) total knee arthroplasty; and (4) intended use of combined spinal epidural anesthesia. The exclusion criteria were as follows: (1) preoperative MMSE score < 24; (2)use of drugs or psychotropic substance abuse, and prolonged use of steroid and hormone medicines; (3) recently undergone major surgical procedures; (4) severely impaired vision and hearing; (5)presurgical abnormalities in the blood; (6) CNS infection; head trauma; multiple sclerosis; neurodegenerative diseases except for Alzheimer’s diseases or other serious nervous conditions; (7) major mental disorders; (8) severe systemic illness which might influence the cerebrospinal fluid(CSF); or Alzheimer-Associated biomarkers including Aβ and tau; and (9) family history of genetic diseases.

Anesthesia and surgery

In PNDABLE, patients received lumbar puncture combined with anesthesia. The anesthesia position is in the transverse decubitus, and the insertion point is situated in the middle of the vertebral column at the L3-L4 level. 2 ml CSF were drawn from the subarachnoid space and 2.5 mL of 0.66% ropivacaine were administered for about 30s. The cerebrospinal fluid sample was centrifuged for 10 minutes at 2,000 x g and stored in an enzyme free EP (Eppendorf) tube (AXYGEN; PCR-02-C) at -80 ° C for further analysis. After anesthesia, sensation levels are controlled below level T8, below chest level. During the procedure, patients were given 5 L/min oxygen flow through masks and moderate vasoactive agents were given to maintain stable vital signs.

Delirium Assessment and sample collection

The Mini-Mental State Examination (MMSE) was used to evaluate the cognition of the participants before surgery by a neurologist [28]. Those who had a MMSE score of less than 24 were excluded from the study. Assessment of delirium was performed twice daily between 9:00 and 10:00 a.m. and between 2:00 and 3:00 p.m., and between 1 and 7 days before discharge from the hospital by an anesthesiologist. POD was defined using Confusion Assessment Method (CAM ) [29], and the severity of POD was measured using the Memorial Delirium Assessment Scale (MDAS) [30, 31].Consequently, postoperative CAM-positive and MDAS-positive patients were documented.

After the spinal-epidural subarachnoid puncture, 2 mL of CSF was taken and put into in a 10 mL polypropylene tubes for two hours. Immediately, all CSF samples were centrifuged at 2000 grams at ambient temperature for 10 minutes, followed by storage at -80°C for further analysis. The concentrations of Aβ42, total Tau (t-Tau), and phosphorylated Tau (p-Tau) in CSF were determined by Enzyme-linked immunosorbent assays (ELISAs). All CSF specimens are randomly placed into one set of plates. In order to avoid any variance in the lot, all antibodies and plates were taken from the same batch. In addition, the coefficients of variation (CVs) within and across lots was below 5% and 15%. All experiments were carried out by investigators who did not know who the patient was.

Sample size estimation

The PNDRFAP and PNDABLE study showed that 8 covariates (age, educational level, BMI, MMSE, albuminuria, duration of surgery, duration of anesthesia and estimated blood loss) were included in the logistic regression. The incidence of postoperative delirium (POD) was 15%, and the loss to follow-up rate was assumed to be 20%. Therefore, the required sample size was calculated to be 667 cases (8 × 10 ÷ 0.15 ÷ 0.8 = 667). A total of 577 patients were enrolled into PNDABLE and 530 patients were enrolled into PNDRF P group (Fig. 1).

Statistical Analysis

The metrological measurement data are verified to follow a normal distribution using the Kolmogorov-Smirnov method. For continuous variables with a normal distribution, the data is expressed as mean standard deviation (SD), while non-normal distributions are represented by median and quartile ranges (M, P25-P75). The independent two-sample t-test was utilized to compare normally distributed data, while the Mann-Whitney U test was employed for non-normally distributed data. Frequency and percentage are reported for categorical data, and chi-square tests are employed for comparisons between groups. The differences in biomarker levels between the postoperative delirium group and the postoperative non-delirium group were depicted in box plots. The difference was considered statistically significant at p < 0.05.

The logistic regression equation was utilized to analyze the relationship between POD and biomarkers, such as total bilirubin, indirect bilirubin, and direct bilirubin. Model 1 presents the unadjusted result. In order to determine whether this result is affected by confounding factors, the study conducts multiple sensitivity analyses. Model 2 reflects the correction of age, sex, educational level and MMSE. Model 3 corrected for hypertension, drinking history, diabetes, smoking, whereas Model 4 included only participants aged ≥ 65, building on the findings of Model 3. In total, four models were employed to investigate the risk and protection of postoperative delirium (POD).

Furthermore, the outcomes of the Hosmer-Lemeshow goodness-of-fit test are depicted as calibration plot curves, which are utilized to assess logistic regression models. The Receiver Operating Characteristic Curve (ROC) region is used to estimate the accuracy of POD biomarkers diagnostic accuracy. Furthermore, the sensitivity, specificity, and area under the curve (AUC) were also calculated. However, the ROC failed to consider the clinical impact of the model. Therefore, Decision Curve Analysis (DCA) was conducted to further evaluate whether the model had effects that supported clinical decision-making [32]. At the same time, a mediating effect model was used to analyze the process and mechanism of the impact of bilirubin on postoperative cognitive impairment. According to the magnitude of the regression coefficient for each biomarker in the logistic regression model for postoperative cognitive impairment, a value level was assigned to each biomarker. The predicted value of postoperative delirium was then calculated based on the total score.

Additionally, we examined the relationships between bilirubin and CSF POD biomarkers by assessing the mediating effect. Meaning was measured with 10,000 bootstrapping iterations, with the model be checked for age, gender, educational level, and MMSE. Then, we utilized the R project software (version4.1.3; http://www.r-project.org/) to develop a dynamic nomogram based on the results of the multivariate logistic regression model. By analyzing the scores of each influencing factor in the regression model, we were able to predict the probability of postoperative delirium through the summation and analysis of the scores.

Finally, a network calculator has been developed based on the predictive model. This calculator allows the input of patient predictors to generate personalized risk estimates. The data were analyzed using IBM SPSS Statistics 27.0.1 (Solvusoft Corporation, Inc., Chicago, Illinois, USA), GraphPad Prism version 9.0 (GraphPad Software, Inc., La Jolla, CA, USA), and R software version 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).

Participants properties

A total of 577 patients in the PNDRFAP cohort were enrolled from 667 subjects (Fig. 1A). There were no significant differences in gender, BMI, drinking history, smoking history, hypertension, diabetes, MMSE score, duration of anesthesia, duration of surgery, estimated volume of infusion, and intraoperative blood loss between the postoperative delirium and non-postoperative delirium groups (Table 1). For the PNDABLE cohort, we screened 667 participants, of whom 621 finally met the study criteria and were included (Fig. 1B). The two groups, postoperative delirium and non-postoperative delirium, were compared using basic clinical data such as gender, BMI, smoking history, MMSE score, duration of anesthesia, estimated volume of infusion, intraoperative blood loss and duration of surgery. The disparity was not statistically significant (Table 2).

Table 1

Characteristics of participants in PNDRFAP
	PNDRFAP
Characteristic	NPOD (n = 474)	POD (n = 103)	P
Age [year, M(Q)]	62(11)	75(8)	< 0.001
Male [n(%)]	277 (58.40)	70 (68.00)	0.074
BMI [kg.m^− 2, M(Q)]	25.00(5.00)	26.00 (5.00)	0.182
Education [year, M(Q)]	12(3)	6(4)	< 0.001
Smoking history [n(%)]	131(27.60)	22 (21.40)	0.191
Drinking history [n(%)]	161(34.00)	26 (25.20)	0.086
Hypertension [n(%)]	228(48.10)	57(55.30)	0.183
Diabetes [n(%)]	95(20.00)	26(25.20)	0.240
MDAS [scores, M(Q)]	1(6)	15(8)	< 0.001
MMSE [scores, M(Q)]	28(3)	28(3)	0.826
Total bilirubin [µmol/L, M(Q)]	8.5(3.30)	13.70(3.60)	< 0.001
Direct bilirubin [µmol/L, M(Q)]	5.60(3.53)	3.40(2.10)	< 0.001
Indirect bilirubin [µmol/L, M(Q)]	2.50(1.33)	10.20(3.80)	< 0.001
Duration of anesthesia (min)	145.20(79.80)	139.80(70.20)	0.615
Duration of surgery (min)	94.80(64.80)	90.00(57.60)	0.896
Estimated volume of infusion (ml)	1100.00(500.00)	1100.00(600.00)	0.058
Intraoperative blood loss (ml)	20.00(195.00)	50.00(190.00)	0.648
Continuous variable use Student's t test or Mann-Whitney U, Categorical variable use Chi-square test.
The numerical variables of normal distribution are statistically described by Average±Standard deviation [‾x±s].
Non-normally distributed numerical variables are statistically described by Median (Interquartile spacing) [M(Q)].
Categorical variables are statistically described by Sample size (Percent) [n(%)].
^*P < 0.05

Table 2

Characteristics of participants in PNDABLE
	PNDABLE
Characteristic	NPOD (n = 504)	POD (n = 117)	P
Age [year, M(Q)]	61.50(11)	74(9)	< 0.001
Male [n(%)]	291 (57.74)	76 (64.96)	0.152
BMI [kg.m^− 2, M(Q)]	25.32 (4.77)	25.71(4.95)	0.111
Education [year, M(Q)]	12(3)	7(4)	< 0.001
Smoking history [n(%)]	138(27.38)	26 (22.22)	0.254
Drinking history [n(%)]	177 (35.12)	26 (22.22)	0.007
Hypertension [n(%)]	169 (33.53)	54(46.15)	0.010
Diabetes [n(%)]	75 (14.88)	31(26.50)	0.003
MDAS [scores, M(Q)]	1(6)	15(7)	< 0.001
MMSE [scores, M(Q)]	28(3)	29(3.5)	0.826
Total bilirubin [µmol/L, M(Q)]	11.90(5.25)	12.40(5.10)	< 0.001
Direct bilirubin [µmol/L, M(Q)]	2.40(1.40)	2.50(1.40)	< 0.001
Indirect bilirubin [µmol/L, M(Q)]	9.40(4.30)	9.70(4.50)	< 0.001
Aβ42 [pg/ml, M(Q)]	357.46(238.64)	288.73(153.00)	< 0.001
P-tau [pg/ml, M(Q)]	39.98(15.79)	45.18 (20.84)	< 0.001
T-tau [pg/ml, M(Q)]	179.48(82.17)	221.55(165.71)	< 0.001
Aβ₄₂/P-tau [M(Q)]	8.85(5.60)	6.34(4.31)	< 0.001
Aβ₄₂/T-tau [M(Q)]	1.97(1.26)	1.31(1.01)	< 0.001
Duration of anesthesia (min)	150.00(79.80)	145.20(65.70)	0.522
Duration of surgery (min)	100.20(70.20)	90.00(62.70)	0.973
Estimated volume of infusion (ml)	1100.00(600.00)	1100.00(500.00)	0.288
Intraoperative blood loss (ml)	20.00(195.00)	50.00(190.00)	0.438
PNDABLE, the Perioperative Neurocognitive Disorder and Biomarker Lifestyle study.
POD, postoperative delirium; NPOD, non-postoperative delirium;
Aβ42, amyloid beta42; t-Tau, total Tau; p-Tau, phosphorylated Tau;
Continuous variable use Student's t test or Mann-Whitney U, Categorical variable use Chi-square test.
The numerical variables of normal distribution are statistically described by Average±Standard deviation [‾x±s].
Non-normally distributed numerical variables are statistically described by Median (Interquartile spacing) [M(Q)].
Categorical variables are statistically described by Sample size (Percent) [n(%)].
^*P < 0.05

In the PNDRFAP study, it was found that there were significant differences in the levels of total bilirubin, direct bilirubin, and indirect bilirubin between POD patients and NPOD patients (Total bilirubin, OR (95% CI): 1.839(1.645–2.056) P < 0.001; Direct bilirubin, OR (95% CI): 1.725(1.448–2.054) P < 0.001; Indirect bilirubin, OR (95% CI): 1.558(1.424–1.705) P < 0.001) (Table 3). In the PNDABLE cohort, the levels of total bilirubin, direct bilirubin, indirect bilirubin, Aβ42, tau, and P-tau were also found to be higher in the postoperative delirium group compared to those in the non-postoperative delirium group. Other variations in the data are illustrated in Fig. 2. We compared bilirubin markers and biomarkers in cerebrospinal fluid between patients with postoperative delirium (POD) and those without postoperative delirium (NPOD). The PNDABLE cohort study found that preoperative levels of total bilirubin, indirect bilirubin, direct bilirubin, P-tau, and T-tau were higher in patients with postoperative delirium (POD) compared to those without postoperative delirium (NPOD). [Total bilirubin, OR (95% CI): 1.203 (1.137–1.274), P < 0.001; Direct bilirubin, OR (95% CI): 1.724 (1.465–2.030), P < 0.001; Indirect bilirubin, OR (95% CI): 1.182 (1.109–1.260), P < 0.001; P-tau, OR (95% CI): 1.042 (1.025–1.060), P < 0.001; T-tau, OR (95% CI): 1.008 (1.006–1.011), P < 0.001] (Table 4).

Table 3

Logistic regression on analysis and sensitivity analysis in PNDRFAP study
	Model 1^a		Model 2^b		Model 3^c		Model 4^d
	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Total bilirubin, µmol/L	1.839(1.645–2.056)	< 0.001	1.673(1.434–1.952)	< 0.001	1.680(1.430–1.973)	< 0.001	1.686(1.424–1.995)	< 0.001
Direct bilirubin, µmol/L	1.725(1.448–2.054)	< 0.001	1.773(1.261–2.494)	< 0.001	1.824(1.281–2.595)	< 0.001	1.856(1.264–2.725)	0.002
Indirect bilirubin, µmol/L	1.558(1.424–1.705)	< 0.001	1.494(1.312–1.702)	< 0.001	1.504(1.310–1.727)	< 0.001	1.520(1.314–1.758)	< 0.001
Note: OR = odds ratio; 95%CI = 95% confidence interval; Aβ₄₂ = β-amyloid42; T-tau = total tau; P-tau = phosphorylated tau
^aModel 1: Unadjusted
^bModel 2: Adjusted for age(50–90), sex, years of education and MMSE scores
^cModel 3: Adjusted for age(50–90), sex, years of education, MMSE scores, smoking history, drinking history, hypertension and diabetes
^dModel 4: Adjusted for age ≥ 65, sex, years of education, MMSE scores, smoking history, drinking history, hypertension and diabetes
^*P < 0.05

Table 4

Logistic regression on analysis and sensitivity analysis in PNDABLE study
	Model 1^a		Model 2^b		Model 3^c		Model 4^d
	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P	OR (95% CI)	P
Total bilirubin, µmol/L	1.203(1.137–1.274)	< 0.001	1.208(1.111–1.312)	< 0.001	1.210(1.110–1.318)	< 0.001	1.168(1.064–1.284)	< 0.001
Direct bilirubin, µmol/L	1.724(1.465–2.030)	< 0.001	1.602(1.284–1.999)	< 0.001	1.772(1.395–2.251)	< 0.001	1.641(1.267–2.125)	< 0.001
Indirect bilirubin, µmol/L	1.182(1.109–1.260)	< 0.001	1.2180(1.075–1.294)	< 0.001	1.170(1.064–1.286)	0.001	1.127(1.014–1.253)	0.026
Aβ₄₂, pg/ml	0.996(0.995–0.998)	< 0.001	0.995(0.993–0.997)	< 0.001	0.995(0.993–0.997)	< 0.001	0.994(0.992–0.997)	< 0.001
P-tau, pg/ml	1.042(1.025–1.060)	< 0.001	1.030(1.008–1.053)	0.008	1.031(1.007–1.056)	0.010	1.027(1.0001–1.1055)	0.046
T-tau, pg/ml	1.008(1.006–1.011)	< 0.001	1.006(1.003–1.010)	< 0.001	1.007(1.003–1.010)	< 0.001	1.005(1.001–1.009)	0.011
Aβ₄₂/P-tau	0.834(0.783–0.889)	< 0.001	0.827(0.762–0.898)	< 0.001	0.824(0.756–0.899)	< 0.001	0.809(0.731–0.896)	< 0.001
Aβ₄₂/T-tau	0.390(0.293–0.519)	< 0.001	0.434(0.303–0.621)	< 0.001	0.418(0.288–0.606)	< 0.001	0.442(0.293–0.666)	< 0.001
Note: OR = odds ratio; 95%CI = 95% confidence interval; Aβ₄₂ = β-amyloid42; T-tau = total tau; P-tau = phosphorylated tau
^aModel 1: Unadjusted
^bModel 2: Adjusted for age(50–90), sex, years of education and MMSE scores
^cModel 3: Adjusted for age(50–90), sex, years of education, MMSE scores, smoking history, drinking history, hypertension and diabetes
^dModel 4: Adjusted for age ≥ 65, sex, years of education, MMSE scores, smoking history, drinking history, hypertension and diabetes
^*P < 0.05

Sensitivity analysis

The PNDRFAP (Table 3) and PNDABLE (Table 4) studies were analyzed in a sensitive manner with three models, namely Model 2, Model 3, and Model 4, in order to verify their reliability.

Adjustment for age, gender, educational level, and MMSE (model 2), there was significant correlation between total bilirubin, indirect bilirubin, direct bilirubin, P-tau, T-tau and the occurrence of POD based on a binary logistic regression model [Total bilirubin, OR (95% CI) 1.208 (1.111–1.312), P < 0.001; Direct bilirubin, OR (95% CI) 1.602 (1.284–1.999), P < 0.001; Indirect bilirubin, OR (95% CI) 1.218 (1.075–1.294), P < 0.001]. The results remained consistent in Model 3 after adjusting for hypertension, diabetes, smoking history, and alcohol consumption history. [Total bilirubin, OR (95% CI) 1.210 (1.110–1.318), P < 0.001; Direct bilirubin, OR (95% CI) 1.772 (1.395–2.251), P < 0.001; Indirect bilirubin, OR (95% CI) 1.170 (1.064–1.286), P < 0.001]. Based on Model 3, the age of the patients was controlled, and the results were consistent. [Total bilirubin, OR (95%CI)1.168(1.064–1.284) P < 0.001; Direct bilirubin, OR (95%CI)1.641(1.267–2.125) P < 0.001; Indirect bilirubin, OR (95%CI)1.127(1.014–1.253) P < 0.026] (Table 3). In PNDRFAP studies, the levels of total bilirubin, direct bilirubin, and indirect bilirubin were compared between the POD and NPOD groups. The results showed that the levels of total bilirubin, direct bilirubin, and indirect bilirubin were significantly higher in the POD group compared to the NPOD group [Total bilirubin, OR (95% CI) 1.839 (1.645–2.056), P < 0.001; Direct bilirubin, OR (95% CI) 1.725 (1.448–2.054), P < 0.001; Indirect bilirubin, OR (95% CI) 1.558 (1.424–1.705), P < 0.001]. These findings were consistent across Model 2, Model 3, and Model 4.

Calibration curve and nomogram

A calibration curve of the incidence of postoperative delirium (POD) in the PNDABLE study demonstrated strong agreement between predicted and observed values. Elevated total bilirubin, direct bilirubin, and indirect bilirubin may be the risk factors of postoperative delirium (POD). (Fig. 3A) There is an increased risk for postoperative delirium (POD) within certain ranges along with increases in total bilirubin, direct bilirubin and indirect bilirubin. Aβ42 may be the protective factor of POD after operation. As Aβ42 content decreases, there is an increase in the likelihood of POD development. Furthermore, regression coefficients were used to assign a value level to each factor in the model. The total score was then obtained by adding up each score. Based on the data from this study, the patient's score ranged from 159.58 to 301.59. In the end, the functional transformation relationship between the total score and the probability of postoperative delirium could be calculated. (Fig. 3B)

Receiver operating characteristic curve(ROC)

The ROC curve indicated that total bilirubin, direct bilirubin, and indirect bilirubin all exhibited strong predictive properties for postoperative delirium (total bilirubin AUC = 0.681; direct bilirubin AUC = 0.706; indirect bilirubin AUC = 0.633). Additionally, the combination of total bilirubin, direct bilirubin, and indirect bilirubin was a more accurate predictor of postoperative delirium than a single indicator (AUC = 0.713). If bilirubin markers are combined with biomarkers in cerebrospinal fluid, the prediction of postoperative delirium is more accurate (AUC = 0.812) (Fig. 4).

Decision curve analysis

The net benefit (y-axis) of biological indicators is plotted against various threshold probabilities (x-axis) on the decision curve (Fig. 5). It can be observed in Fig. 5 that total bilirubin, direct bilirubin, indirect bilirubin, and the combination of the three bilirubin indicators show differences in net benefit. The combination of multiple indicators provides a greater net benefit compared to total bilirubin, direct bilirubin, and indirect bilirubin, as well as in comparison to the clinical default strategies of "treat all" or "treat none". Therefore, compared to a single index, the combination of multiple indicators can better discriminate and guide clinical diagnosis and treatment.

Causal mediation analyses

Binary logistic regression in PNDABLE studies revealed that total bilirubin, direct bilirubin, indirect bilirubin, T-tau, P-tau, Aβ42/P-tau, and Aβ42/tau were positively correlated with POD, while Aβ42 was negatively correlated with POD. Therefore, total bilirubin, direct bilirubin, and indirect bilirubin, as well as T-tau, P-tau, Aβ42/P-tau, and Aβ42/tau, are all risk factors for POD, while Aβ42 is a protective factor for POD. We are aware that Aβ42, T-tau, P-tau, Aβ42/P-tau, and Aβ42/tau have a specific impact on POD. Therefore, we are further investigating whether total bilirubin, direct bilirubin, and indirect bilirubin can mediate the effects of T-tau, P-tau, Aβ42/P-tau, and Aβ42/tau on POD. Mediation analysis revealed that the association between total bilirubin and POD was mediated by T-tau, P-tau, Aβ42/P-tau, and Aβ42/tau. Similarly, the connection between direct bilirubin, indirect bilirubin, and POD was also mediated by T-tau, P-tau, Aβ42/P-tau, and Aβ42/tau (Fig. 6). The study revealed that the association between total bilirubin, direct bilirubin, indirect bilirubin, and postoperative delirium (POD) was influenced by CSF t-tau, with respective proportions of 9.08%, 10.26%, and 8.70%.

Prediction model development

In order to enhance clinical diagnosis and treatment, a dynamic nomogram is created. We developed a dynamic nomogram model to forecast the likelihood of postoperative delirium (POD)occurrence using eight independent predictors, which include Aβ42, T-tau, P-tau, Aβ42/tau, total bilirubin, direct bilirubin, and indirect bilirubin. Users can access the web page (https://bylqdsslyy.shinyapps.io/dynnomapp/), input variables on the page, and then click "Predict" in the lower left corner of the page to obtain the probability of POD occurring. In addition, the dynamic nomogram also provides a graphical summary, numerical summary, and model summary. (Fig. 7)

We studied the relationship of bilirubin with postoperative delirium (POD) in this study. The study showed total bilirubin, direct bilirubin, and indirect bilirubin were identified as risk factors for postoperative delirium. Furthermore, we found that these risk factors may be linked to specific biomarkers in the cerebrospinal fluid, such as Aβ42, T-tau, and P-tau.

Recent studies have shown that postoperative delirium is the most common surgical complication in the elderly, with an incidence rate ranging from 5–50% of elderly patients. The main clinical manifestations include inattention, emotional changes, confusion, and alterations in consciousness [33, 34]. Nowadays, as the population ages, the number of surgeries is increasing annually, leading to a rise in the incidence of postoperative delirium. The causes of postoperative delirium are extensive and complex, and the pathophysiological mechanism is not clear. The existing evidence supports theories such as neuroinflammatory response [35] and neurotransmitter disorders [36], but there is still a lack of conclusive evidence. The search for biomarkers related to postoperative delirium can not only provide evidence of the pathophysiological mechanism but also contribute to the prediction and treatment of postoperative delirium.

We found that total bilirubin, direct bilirubin, and indirect bilirubin are risk factors for POD, and we validated the independence and accuracy of the results using multiple models.

Previous studies have shown that bilirubin is a risk factor for the nervous system. Bilirubin triggers the activation of astrocytes and microglia, resulting in the increased expression of pro-inflammatory cytokines, such as tumor necrosis factor-α, Interleukin-1β, and IL-6, which in turn leads to inflammation and necrosis of nerve cells. Bilirubin can disrupt the energy metabolism of brain cell mitochondria, leading to demyelination of glial cells, reduction in nerve cell membrane potential, and impact on synaptic transmission. Ultimately, these effects can lead to changes in neuronal morphology and function [37–39]. A recent study also reached the same conclusion, indicating that the incidence of POD is positively correlated with total bilirubin when the level of total bilirubin is below 20 mg/dL [40]. There is a study that has shown that patients with postoperative delirium have significantly higher preoperative mean bilirubin levels than non-delirium patients[41].

However, it has been shown that bilirubin have a protective effect on the nervous system [21]. For example, Doré et al. demonstrated that low concentrations of bilirubin can mitigate the toxicity of hydrogen peroxide in rat hippocampal and cortical neuronal cultures. At higher concentrations of bilirubin, the protective effect of bilirubin on hippocampal neurons treated with hydrogen peroxide diminishes [42]. This contradicts the results of our study, which may be attributed to the small sample size and the serendipitous nature of the findings. At the same time, they may focus on the antioxidant effects of bilirubin and overlook its toxicity to the nervous system.

Existing studies have shown that patients with preoperative cognitive impairment are more likely to experience postoperative delirium[43]. Postoperative delirium and postoperative cognitive impairment share similarities in pathological and physiological mechanisms, and some biomarkers in cerebrospinal fluid also overlap. Ashok et al.'s study demonstrated that levels of Aβ 42, T-tau, and P-tau in cerebrospinal fluid are associated with cognitive dysfunction [44–46], Additionally, another study found that these biological markers are also linked to the occurrence of postoperative delirium [47]. Our research results also support this viewpoint. Meanwhile, we found that these biomarkers partially mediate the effect of bilirubin on postoperative cognitive impairment.

It was found that the T-tau protein was primarily distributed in the frontal lobe, temporal lobe, hippocampus, and neuronal axons in the entorhinal region of the brain [48]. Bilirubin, on the other hand, frequently leads to necrosis of the basal ganglia, hippocampus, inferior colliculus, and cerebellar neurons. Dal Ben et al. found that hippocampal neurons were severely damaged in the brain tissue of neonatal rats with bilirubin-induced brain injury [49]. After bilirubin crosses the blood-brain barrier, it triggers oxidative stress, endoplasmic reticulum stress, and inflammatory responses. This leads to alterations in mitochondrial and endoplasmic reticulum membranes, and activates signaling pathways such as p38MAPK to initiate the caspase apoptosis pathway. The T-tau protein is then cleaved as a substrate for caspase-3, and the cleaved T-tau fragment promotes the expression of caspase-3 [50–52]. When the caspase-3 gene protein is expressed in the hippocampus, the surrounding nerve cells undergo apoptosis and necrosis, resulting in a decrease in nerve cell density and disruption of cognitive learning function in the hippocampus [53–55].

In this study, there are also several limitations. Firstly, this experiment only utilized data from nearly 1000 individuals in two databases for analysis. The small number and scale of data samples still limit the analysis. Expanding the sample size can lead to more comprehensive research, and the results will be more convincing. Second, this study focused on the correlation between bilirubin levels and postoperative delirium in patients undergoing surgical surgery. However, the specific mechanism of postoperative delirium related to bilirubin is not well understood, and there is no theoretical basis to support it. This can be supplemented with additional basic animal experiments. Finally, this study exclusively examines the correlation between biological indicators in cerebrospinal fluid and bilirubin, and does not investigate the relationship between other indicators that may impact postoperative delirium and bilirubin.

In summary, elevated total bilirubin, direct bilirubin, and indirect bilirubin are preoperative risk factors that contribute to the development of postoperative delirium (POD). The higher the concentration of bilirubin, the greater the likelihood of experiencing postoperative delirium. The association between bilirubin and postoperative delirium may be partly linked to specific biomarkers in cerebrospinal fluid.

Acknowledgements

The authors thank the colleagues who have made contributions to build the PNDABLE Study and PNDRFAP. The authors also express their gratitude to the participants and their families for their cooperation in this study.

Authors^,contributions

Y.L.B. conceived the current study. B.W., X.L., R.D., H.Y.G., Y.L.W., S.L.X., and J.K. performed the experiments. S.H.H., Y.N.L. and C.L. analyzed data. S.H.H., S.L.X. and B.W. performed the experiments, wrote and revised the manuscript. All authors have contributed to the manuscript revising and editing critically for important intellectual content and given final approval of the version and agreed to be accountable for all aspects of the work presented here. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (91849126).

Data and materials availability

Data sets used and/or analyzed in a current study shall, upon reasonable request, be made available by the correspondent author.

Ethical Approval and Participation Consent

The study was carried out according to the Helsinki Declaration of Helsinki Principles. Qingdao Municipal Hospital’s Ethics Committee approved this study (PNDABLE: ChiCTR2000033439, registration date: 01/06/2020, PNDRAFP: ChiCTR2000033639, registration date: 07/06/2020). The patients/participants gave their written informed consent to participate in this study.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests with respect to the research, authorship, and/or publication of this article.

Inouye SK, Westendorp RG, Saczynski JS, et al. Delirium in elderly people–authors'reply. Lancet. 2014;383(9934):2045.
Oh ST, Park JY. Postoperative delirium. Korean J Anesthesiol. 2019;72(1):4–12.
Schenning KJ, Deiner SG. Postoperative Delirium in the Geriatric Patient. Anesthesiol Clin. 2015;33(3):505–16.
Goldberg TE, Chen C, Wang Y, et al. Association of Delirium With Long-term Cognitive Decline: A Meta-analysis. JAMA Neurol. 2020;77(11):1373–81.
Maldonado JR. Neuropathogenesis of delirium: review of current etiologic theories and common pathways. Am J Geriatr Psychiatry. 2013;21(12):1190–222.
Nho K, Kueider-Paisley A, MahmoudianDehkordi S, et al. Altered bile acid profile in mild cognitive impairment and Alzheimer's disease: Relationship to neuroimaging and CSF biomarkers. Alzheimers Dement. 2019;15(2):232–44.
Dillon ST, Vasunilashorn SM, Otu HH, et al. Aptamer-Based Proteomics Measuring Preoperative Cerebrospinal Fluid Protein Alterations Associated with Postoperative Delirium. Biomolecules. 2023;13(9):1395.
Fagan AM, Mintun MA, Shah AR, et al. Cerebrospinal fluid tau and ptau(181) increase with cortical amyloid deposition in cognitively normal individuals: implications for future clinical trials of Alzheimer's disease. EMBO Mol Med. 2009;1(8–9):371–80.
Blennow K, Zetterberg H. Biomarkers for Alzheimer's disease: current status and prospects for the future. J Intern Med. 2018;284(6):643–63.
Tiwari S, Atluri V, Kaushik A, et al. Alzheimer's disease: pathogenesis, diagnostics, and therapeutics. Int J Nanomed. 2019;14(7):5541–54.
Cunningham EL, McGuinness B, McAuley DF, et al. CSF Beta-amyloid 1–42 Concentration Predicts Delirium Following Elective Arthroplasty Surgery in an Observational Cohort Study. Ann Surg. 2019;269(6):1200–05.
Maruhashi T, Kihara Y, Higashi Y. Bilirubin and Endothelial Function. J Atheroscler Thromb. 2019;26(8):688–96.
Thakkar M, Edelenbos J, Dore S. Bilirubin and Ischemic Stroke: Rendering the Current Paradigm to Better Understand the Protective Effects of Bilirubin. Mol Neurobiol. 2019;56(8):5483–96.
Higuchi S, Kabeya Y, Uchida J, et al. Low Bilirubin Levels Indicate a High Risk of Cerebral Deep White Matter Lesions in Apparently Healthy Subjects. Sci Rep. 2018;8(1):6473.
Brito MA, Vaz AR, Silva SL, et al. N-methyl-aspartate receptor and neuronal nitric oxide synthase activation mediate bilirubin-induced neurotoxicity. Mol Med. 2010;16(9–10):372–80.
Feng J, Li M, Wei Q, et al. Unconjugated bilirubin induces pyroptosis in cultured rat cortical astrocytes. J Neuroinflammation. 2018;15(1):23.
Beckmann S, Schubert M, Burkhalter H, et al. Postoperative Delirium After Liver Transplantation is Associated With Increased Length of Stay and Lower Survival in a Prospective Cohort. Prog Transpl. 2017;27(1):23–30.
Wang SH, Wang JY, Lin PY, et al. Predisposing risk factors for delirium in living donor liver transplantation patients in intensive care units. PLoS ONE. 2014;9(5):e96676.
Lescot T, Karvellas CJ, Chaudhury P, et al. Postoperative delirium in the intensive care unit predicts worse outcomes in liver transplant recipients. Can J Gastroenterol. 2013;27(4):207–12.
Yoon JS, Kim YR, Choi JW, et al. Risk factors of postoperative delirium following liver transplantation. Korean J Anesthesiol. 2009;57(5):584–89.
Kim TS, Pae CU, Yoon SJ, et al. Decreased plasma antioxidants in patients with Alzheimer's disease. Int J Geriatr Psychiatry. 2006;21(4):344–8.
Yin XL, Jia QF, Zhang GY, et al. Association between decreased serum TBIL concentration and immediate memory impairment in schizophrenia patients. Sci Rep. 2019;9(1):1622.
Brites D. Bilirubin injury to neurons and glial cells: new players, novel targets, and newer insights. Semin Perinatol. 2011;35(3):114–20.
Wang F, Tang X, Wang J, et al. Potential Value of Serum Uric Acid in the Identification of Postoperative Delirium in Geriatric Patients Undergoing Knee Replacement. Front Aging Neurosci. 2022;14:909738.
Deng X, Qin P, Lin Y, et al. The relationship between body mass index and postoperative delirium. Brain Behav. 2022;12(4):e2534.
Wu X, Zhang N, Zhou B, et al. Alcohol consumption may be associated with postoperative delirium in the elderly: the PNDABLE study. BMC Anesthesiol. 2023;23(1):222.
Lin Y, Yu N, Lin X, et al. Preoperative cerebrospinal fluid biomarkers may be associated with postoperative delirium in patients undergoing knee/hip arthroplasty: the PNDABLE study. BMC Geriatr. 2023;23(1):282.
Mitchell AJ, Shukla D, Ajumal HA, et al. The Mini-Mental State Examination as a diagnostic and screening test for delirium: systematic review and meta-analysis. Gen Hosp Psychiatry. 2014;36(6):627–33.
Inouye S, van Dyck C, Alessi C, et al. Clarifying Confusion: The Confusion Assessment Method A New Method for Detection of Delirium. Ann Intern Med. 1990;113(12):941–48.
De J, Wand AP. Delirium Screening: A Systematic Review of Delirium Screening Tools in Hospitalized Patients. Gerontologist. 2015;55(6):1079–99.
Wong CL, Holroyd-Leduc J, Simel DL, et al. Does This Patient Have Delirium? Value of Bedside Instruments. JAMA Neurol. 2012;304(7):779–86.
Van Calster B, Wynants L, Verbeek JFM, et al. Reporting and Interpreting Decision Curve Analysis: A Guide for Investigators. Eur Urol. 2018;74(6):796–804.
Palanca BJA, Wildes TS, Ju YS, et al. Electroencephalography and delirium in the postoperative period. Br J Anaesth. 2017;119(2):294–307.
Zwischenberger JB, Pitcher HT. Extracorporeal Membrane Oxygenation Management: Techniques to Liberate from Extracorporeal Membrane Oxygenation and Manage Post-Intensive Care Unit Issues. Crit Care Clin. 2017;33(4):843–53.
Xie Z, Enkhjargal B, Reis C, et al. Netrin-1 Preserves Blood-Brain Barrier Integrity Through Deleted in Colorectal Cancer/Focal Adhesion Kinase/RhoA Signaling Pathway Following Subarachnoid Hemorrhage in Rats. J Am Heart Assoc. 2017;6(5):e005198.
Maldonado JR. Delirium pathophysiology: An updated hypothesis of the etiology of acute brain failure. Int J Geriatr Psychiatry. 2018;33(11):1428–57.
Deliktas M, Ergin H, Demiray A, et al. Caffeine prevents bilirubin-induced cytotoxicity in cultured newborn rat astrocytes. J Matern Fetal Neonatal Med. 2019;32(11):1813–19.
Watchko JF, Maisels MJ. The enigma of low bilirubin kernicterus in premature infants: why does it still occur, and is it preventable? Semin Perinatol. 2014;38(7):397–406.
Brites D. The evolving landscape of neurotoxicity by unconjugated bilirubin: role of glial cells and inflammation. Front Pharmacol. 2012;3(5):88.
Lu RY, Zhu HK, Liu XY, et al. A Non-Linear Relationship between Preoperative Total Bilirubin Level and Postoperative Delirium Incidence after Liver Transplantation. J Pers Med. 2022;12(2):141.
Park KH, Son HJ, Choi YJ, et al. Liver Transplant Patients with High Preoperative Serum Bilirubin Levels Are at Increased Risk of Postoperative Delirium: A Retrospective Study. J Clin Med. 2020;9(5):1591.
Doré S, Takahashi M, Ferris CD, et al. Bilirubin, formed by activation of heme oxygenase-2, protects neurons against oxidative stress injury. Proc Natl Acad Sci U S A. 1999;96(5):2445–50.
Lee HB, Mears SC, Rosenberg PB, et al. Predisposing factors for postoperative delirium after hip fracture repair in individuals with and without dementia. J Am Geriatr Soc. 2011;59(12):2306–13.
Frolich L, Peters O, Lewczuk P, et al. Incremental value of biomarker combinations to predict progression of mild cognitive impairment to Alzheimer's dementia. Alzheimers Res Ther. 2017;9(1):84.
van Rossum IA, Vos SJ, Burns L, et al. Injury markers predict time to dementia in subjects with MCI and amyloid pathology. Neurology. 2012;79(17):1809–16.
Ashok A, Singh N, Chaudhary S, et al. Retinal Degeneration and Alzheimer's Disease: An Evolving Link. Int J Mol Sci. 2020;21(19):7290.
Ballweg T, White M, Parker M, et al. Association between plasma tau and postoperative delirium incidence and severity: a prospective observational study. Br J Anaesth. 2021;126(2):458–66.
Musi N, Valentine JM, Sickora KR, et al. Tau protein aggregation is associated with cellular senescence in the brain. Aging Cell. 2018;17(6):e12840.
Dal Ben M, Bottin C, Zanconati F, et al. Evaluation of region selective bilirubin-induced brain damage as a basis for a pharmacological treatment. Sci Rep. 2017;7(1):41032.
Bulat N, Widmann C. Caspase substrates and neurodegenerative diseases. Brain Res Bull. 2009;80(4–5):251–67.
Araujo IM, Carreira BP, Pereira T, et al. Changes in calcium dynamics following the reversal of the sodium-calcium exchanger have a key role in AMPA receptor-mediated neurodegeneration via calpain activation in hippocampal neurons. Cell Death Differ. 2007;14(9):1635–46.
Bano D, Ankarcrona M. Beyond the critical point: An overview of excitotoxicity, calcium overload and the downstream consequences. Neurosci Lett. 2018;663:79–85.
GM. C. Caspases: the executioners of apoptosis. Biochem J. 1997; 326(Pt 1):1–16.
Farlow MR, Thompson RE, Wei LJ, et al. Double-Blind, Placebo-Controlled, Phase II Study Assessing Safety, Tolerability, and Efficacy of Bryostatin in the Treatment of Moderately Severe to Severe Alzheimer's Disease. J Alzheimers Dis. 2019;67(2):555–70. A Randomized.
Yon JH, Daniel-Johnson J, Carter LB, et al. Anesthesia induces neuronal cell death in the developing rat brain via the intrinsic and extrinsic apoptotic pathways. Neuroscience. 2005;135(3):815–27.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Preoperative plasma bilirubin may be associated with postoperative delirium in old patients undergoing knee/hip arthroplasty: the PNDRFAP study and the PNDABLE study

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

PNDRAFP

Study registration and data availability

Setting and participants

Anesthesia and surgery

PNDABLE

Study registration and data availability

Setting and Participants

Anesthesia and surgery

Delirium Assessment and sample collection

Sample size estimation

Statistical Analysis

Results

Participants properties

Sensitivity analysis

Calibration curve and nomogram

Receiver operating characteristic curve(ROC)

Decision curve analysis

Causal mediation analyses

Prediction model development

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1